More suicide attempts are made during adolescence than at any other time in life. However, the measurement of suicidal ideation in adolescence has been hampered by a failure to understand the fine-grained aspects of suicidal ideation that predict the transition from suicidal thoughts to actual suicidal behavior. Recent research suggests that knowing how adolescents think about suicide ? i.e., the form of suicidal ideation ? in terms of characteristics such as frequency, duration, and pattern ? may help predict not only whether an adolescent will go on to make a future attempt but also how soon they will do so, regardless of whether they have previously made a suicide attempt. However, current suicidal ideation measures and conceptualizations of suicide risk tend to ignore the form of an adolescent's suicidal ideation. This proposal for a renewal of a SCORE SC1 grant seeks to test 1) whether suicidal ideation that is characterized as occurring within a more persistent, long-standing episode is associated with higher short-term risk for a future suicide attempt over a 12-month follow-up period, compared to suicidal ideation that occurs in the context of a brief episode; 2) whether chronic suicidal ideation is characterized by more interference from and difficulty disengaging from suicide-related thoughts, as assessed by laboratory measures; and 3) whether this interference from and difficulty disengaging from suicide-related thoughts explains the greater risk of future suicide attempts among persistent vs. brief suicide ideators. These aims will be addressed via a longitudinal study of a diverse sample of 150 adolescents, ages 12-19, recruited from two New York City hospitals. Adolescents will be interviewed about the circumstances leading to their suicidal ideation and attempts using a newly developed semi-structured interview. They will also complete laboratory measures of their ability to disengage from suicide-related thoughts. Adolescents (and their parents) will be interviewed again 3 and 12 months later to assess for suicidal ideation and attempts made within the follow-up interval. This research will combine a newly developed semi- structured suicidal ideation interview with structured diagnostic interviews, tests of intellectual functioning, self- report assessments of cognitive biases, and objective measures of suicide-related cognitive biases. Data will be analyzed using multilevel modeling. Understanding patterns of suicidal ideation and their underlying cognitive profiles is a critical step in identifying targets for intervention to prevent suicide in this population. This research has the potential to change how suicide risk assessments are conducted in pediatric EDs by identifying which characteristics of suicidal ideation merit focus in clinical assessments with adolescents.